Nutritional Intake Data

This dataset is from the Food and Agriculture Organization (FAO) of the United Nations and contains information on a country’s food supply. It provides data on various food items—covering both primary and processed commodities that are potentially available for human consumption—and details their sources and utilization.

The dataset contains 32,910 rows and 15 columns and is formatted in long format. It spans the period from 2010 to 2021. The key variables used for our analysis are:

• Country: The country from which the data is collected.
• Element: The macronutrient value in grams or kilocalories per capita per day.
• Item: The type of food item, specifying whether it is a vegetable, meat, or other source.
• Year of Consumption: The year for which the data is recorded.

Life Expectancy Data

The dataset titled “Life Expectancy from 2000 to 2021” from the World health Organisation provides data on life expectancy trends across countries. It contains 24420 rows and 34 columns and spans the years from 2000 to 2021. The data is structured in a long format.

Key variables that we used from this dataset:

• Country: The country for which life expectancy data is recorded.
• Year: The year corresponding to the life expectancy data.
• Life Expectancy: The average life expectancy at birth for each country in years.

Geographical Data:

The third dataset includes the geographical coordinates of the capital cities of the countries, which was used to map and visualize the location of each country in relation to the life expectancy and nutritional intake data.

Data Cleaning

Handling Missing Values: We identified and addressed missing values in the datasets. Fortunately, there were no missing values in the critical columns that were necessary for our analysis.

Dropping Unnecessary Variables: Any variables not relevant to our analysis were removed to streamline the data and focus on the essential components.

Data Transformation

The data was transformed to ensure consistency across all datasets. For example, variables such as nutritional intake were normalized to ensure comparability between countries with varying population sizes.

Data Merging:

The three datasets were merged using the country variable as the key. This allowed for the integration of life expectancy, nutritional intake, and geographic information into a single dataset.

Data Manipulation

For our analysis, we categorized countries into three groups based on life expectancy:

• Top 10 countries with the highest life expectancy.
• Middle 10 countries with life expectancy in the middle range.
• Bottom 10 countries with the lowest life expectancy.

Within these groups, we focused on analyzing the highest caloric intake for each year available in the dataset. Additionally, for life expectancy, we selected the most recent and highest value for each country to ensure that our analysis reflected the best available data for each country over the period of study.

Data Analysis

We analyzed the nutritional intake for each country group over time, focusing on key macronutrients such as fat, protein, and carbohydrates. This helped identify trends in dietary intake that may correlate with life expectancy. Further analysis included more specific macronutrients, such as Animal vs. vegetal intake, Protein sources.

Visualizations were created using mostly the ggplot2 package in R, while ensuring that all plots were color-blind safe. Whenever possible, interactive visualizations were produced to enhance user interaction and provide additional insights.

As part of our analysis, we ran a simple linear regression model to analyze the trend in global life expectancy over time. In this model, life expectancy was used as the dependent variable, while the year was set as the independent variable. This approach allowed us to examine whether there was a consistent increase in life expectancy from 2010 to 2021.

Limitations of our approach

With these datasets come several limitations that need to be acknowledged in our analysis. First, it is important to recognize that the nutritional intake data for every individual in a country cannot be perfectly captured. As a result, the quality and accuracy of the data may vary across countries, and data collection in certain regions may be more challenging due to accessibility issues.

Furthermore, there is some uncertainty regarding how the data was collected. It is unclear whether macronutrient intakes were self-reported by individuals or gathered by the respective organizations through other means, which could affect the reliability of the data.

Within countries, there may also be significant regional differences in nutritional intake and life expectancy, which are not accounted for in this analysis. By aggregating countries into top, middle, and bottom categories, we risk oversimplifying complex factors. This approach might mask important nuances in dietary habits and life expectancy trends that are specific to individual regions or countries.

Additionally, while this report primarily focuses on the role of nutrition, it is essential to note that non-dietary factors—such as healthcare access, education, and socio-economic conditions—also play a significant role in influencing life expectancy (as discussed in the Theoretical Focus section).

Use of Generative AI

We used ChatGPT as our Generative AI tool during our work on R tasks, which presented both opportunities and challenges. One challenge we faced was handling data effectively, especially when we wanted to use specific methods introduced during the course, such as data wrangling with dplyr or visualizations with ggplot2. While it was up to us to decide how to approach the tasks, we preferred to follow these methods to maintain consistency with what we had learned. However, when we asked ChatGPT for assistance, it sometimes suggested alternative approaches using different packages or methods, which did not always align with our chosen way of working. This required us to manually adjust the AI’s suggestions to match our goals.

An important aspect of using ChatGPT was the need to clearly and explicitly state our objectives. For example, when we asked for help with data preprocessing, ChatGPT often didn’t fully grasp the context or the specific transformations we needed. As a result, it occasionally generated solutions that weren’t quite suited to our dataset or analysis. We had to carefully communicate exactly what we wanted to achieve and refine the questions we posed to get more relevant output.

That said, ChatGPT proved useful for simpler tasks, such as generating code for basic data cleaning or debugging common errors. However, we consistently had to double-check the output to ensure it met our needs, especially when the AI’s suggestions diverged from the methods we wanted to use. This experience highlighted the importance of clear communication and human oversight when integrating AI into the workflow, ensuring that the results were aligned with our project’s specific goals and methods.

Geographical Patterns of Life Expectancy and Nutritional Intake [Chapter of Choice]

Plot 1: Countries by Mean Life Expectancy

Plot 2: World Map of Categorized Countries

Plot 1 illustrates life expectancy for countries in the top 10, middle 10, and bottom 10 categories. The accompanying map (Plot 2) geographically locates countries from these three groups. Hovering over the dots on the map or the bars in the chart reveals details for each country, such as the country name, its average life expectancy, and the average caloric intake.

From the bar chart, we observe that Japan leads in life expectancy, followed closely by South Korea, Spain, and France. Noticeably, the gap in life expectancy between the countries in the top and middle categories is relatively small. There is a more significant decline in life expectancy when we compare the life expectancy of countries from the middle group with the bottom group. The countries in the bottom category are predominantly located in Africa, particularly near the equator or in southern regions. In contrast, the middle group of countries includes nations from Asia, Africa, and Latin America, reflecting a more geographically diverse mix.

The countries from the top category are more dispersed, covering Europe, the Middle East, East Asia, and Oceania. Upon first inspection, it becomes evident that the countries in the top category are primarily from the developed world, while the middle category includes less developed countries, and the bottom category comprises the least developed nations.

In this plot we can see that life expectancy on a global level has been more or less consistently rising in the past years. Although the COVID-19 pandemic did affect life expectancy in most countries, we treated those values as outliers, as this data cannot be concluded to be a consistent factor in longevity on a worldwide level.

The summary of the model shows us that the rate of increase is about 0.11138 years of life expectancy every year that passes, if analyzed as a linear trend, with a non-insignificant p-value. This means that, on average, every year that goes by, humanity is expected to live around one month and 10 days longer than they were previously.

Overall Caloric Intake and its Relationship with Life Expectancy

Plot 4: Change in Caloric Intake 2010-2021

Next, when considering caloric intake, we see that countries in the top 10 life expectancy group also have the highest caloric intake, followed by the middle 10 countries, with the bottom 10 countries ingesting the least (plot 4). From this, a possible initial relationship emerges: countries with higher life expectancy also tend to consume more calories.

However, when examined more closely, it becomes evident that the caloric intake for the top and middle countries is rising steadily, but at different rates. The middle countries show a sharper increase in caloric intake, which may reflect their improving economic and healthcare conditions. In contrast, the bottom 10 countries show a slight decline in caloric intake over the years, even as their life expectancy rises. This counterintuitive trend suggests that while caloric intake plays a role in life expectancy, other significant factors must also be at play.

From these initial observations, we can condense three main points:

• Countries in the top 10 life expectancy group have had a relatively stable life expectancy since 2010, with only a slight drop during the COVID-19 pandemic. These countries also consume the most calories across the observed period.

• The middle group, while still developing, shows progress: both life expectancy and caloric intake have risen since 2010. Their caloric intake is the second highest among the three groups, and their life expectancy is gradually approaching the levels seen in the top category.

• The bottom 10 countries also show an increase in life expectancy over the years, but their caloric intake has declined slightly. Despite consuming fewer calories, these countries have managed to improve their life expectancy, indicating that factors beyond nutrition—such as healthcare access and disease prevention—could be contributing to these gains.

The Influence of Macronutrient Consumption on Life Expectancy

Plot 5: Nutritional Profile Comparison

With this initial observation, we can consider the second research question and ask, if the consumption of different macronutrients might impact life expectancy. Plot 5 illustrates the nutritional profile for each life expectancy category, displaying the average intake of fat, protein, and total caloric intake per day from 2010 to 2021.

As expected, countries with the highest life expectancy also have the highest overall caloric intake, followed by the middle category, with the bottom category lagging behind. The top and middle countries have shown a steady increase in fat consumption over the years. Also, in the top and middle groups, caloric intake is either gradually increasing or remaining relatively stable. However, the bottom category shows a decline in overall caloric intake, while fat and protein intake – important nutritional components - remain more or less constant.

At this stage, it becomes difficult to draw a direct relationship between life expectancy and caloric intake. While the data for the top and middle categories might suggest that higher food intake leads to higher life expectancy, the bottom group raises questions about the impact of other factors. For the bottom category, where food intake (particularly total calories) has decreased, life expectancy continues to rise (see plot 3). This leads to speculation about whether factors such as healthcare, hygiene, or other social determinants are playing a more significant role than nutrition alone.

Caloric Intake and the Complexities of Life Expectancy

Plot 6: Average intake across different countries

While the previous plots might lead to the assumption that higher caloric intake is tied to a higher life expectancy, Plot 6 demonstrates that this is not necessarily the case. This plot provides a detailed view of all the countries under focus, showing their average caloric intake, broken down into fat, protein, and total calories.

In this plot, the more saturated bars represent the countries from the top 10 life expectancy group, while the lighter bars correspond to countries in the middle and bottom groups. Countries from the top 10 are more concentrated on the left side of the plot, confirming that, on average, they tend to consume more calories. However, an interesting anomaly is Japan, the country with the highest life expectancy, which ranks only 17th out of the 30 countries in terms of daily caloric intake.

Similarly, Lesotho, which has the lowest life expectancy, does not have the lowest caloric intake, further suggesting that there are other contributing factors beyond daily caloric consumption that influence life expectancy. This emphasizes the need to consider other variables such as healthcare quality, access to clean water, and disease prevalence when analyzing the relationship between nutrition and longevity.

Thus, while caloric intake appears to play some role in life expectancy, this plot highlights the complexities of nutrition and its varied effects across countries, where high caloric intake does not always correspond to a longer life.

Animal and Vegetal-Based Product Consumption Across Life Expectancy Categories

Plot 7: Average Daily Calories Consumed from Animal vs Vegetal Products

Plot 7 provides a closer look of the proportion of calories consumed from animal and plant-based products across the three life expectancy categories. The results in the plot align with previous observations that countries in the top life expectancy category consume the most calories, but reveals that the highest proportion of those calories come from animal products.

When looking at the top category, it is noticeable that the intake of both animal and plant-based products has remained relatively stable over time, although there has been a slight rise in animal product consumption. The middle category, however, shows a more significant increase in the consumption of both animal and plant-based products over recent years. This increase might be tied to economic growth and rising wealth among the populations in these countries. As animal products tend to be more expensive, it is probable that as incomes rise, people can afford to incorporate more of these foods into their diets. Another factor might be the increased accessibility of these products due to globalization, which has made food sources from across the world more available to these countries.

The bottom category presents a contrasting pattern, with a slight drop in plant-based product consumption and stable consumption of animal products over the years. This stability in animal product consumption, despite the drop in overall caloric intake, could suggest that while access to food has diminished in quantity, the quality or type of available food has remained constant.

The lack of significant growth in their consumption patterns could highlight the disparity between richer and poorer nations. Although we cannot see individual countries in this plot, the general trend suggests that these countries might be struggling with the economic and infrastructure challenges that restrict their access to diverse food sources.

Animal vs. Vegetal Product Ratios and Regional Dietary Patterns

Plot 8: Ratio of Animal vs. Vegetal Products

Plot 8 provides a more detailed breakdown of the ratio between animal and vegetal products consumed in each country. A lower ratio indicates that the consumption of animal and vegetal products is relatively equal, while a higher ratio suggests a greater difference between the two, with either animal or vegetal products dominating the diet.

The countries in the top category show the lowest and most stable ratios, indicating that their diets are relatively balanced in terms of animal and vegetal product consumption. This balance has been maintained consistently over the years across all top-category countries. In the mid-category, the ratio has gradually dropped over time, indicating a trend toward a more balanced consumption of animal and plant products. Bulgaria stands out as an outlier in the middle category, where a significant drop in the ratio occurred between 2020 and 2021.

The bottom category reveals the most intriguing developments in the ratio of animal to vegetal product consumption. The spread of ratios is the greatest here, with some countries, like Somalia, exhibiting low ratios, suggesting their diet is more similar to those of the top 10 countries. On the other hand, Mozambique shows a much higher ratio, indicating a diet more heavily reliant on plant-based foods compared to animal products.

These differences in the ratios across countries may reflect various regional dietary patterns. While it might be expected that different regions have varied access to or preferences for animal and plant products, this is not the case across all categories. For example, countries from both the top and middle categories are geographically diverse (e.g. Switzerland in Europa and Japan in Asia), yet their ratios remain relatively stable. In contrast, all the bottom-category countries are in Africa, yet their ratios differ widely, suggesting that factors beyond region, such as economic access or cultural preferences, might influence these dietary patterns.

Additionally, the development of these ratios over time is more volatile and less linear in the bottom category, hinting that these countries may experience shifts in their diets more frequently, possibly due to changing economic conditions or varying food availability.

These findings underscore the complexity of food consumption patterns across different countries and suggest that changing eating habits, whether due to economic shifts, access to resources, or cultural influences, play a significant role in shaping these ratios.

Protein Consumption Patterns: Animal vs. Plant Sources

Plot 9: Animal vs. Plant Protein

Plot 9 takes a deeper look at protein consumption by breaking down the sources—whether animal or vegetal. It highlights the part of protein sources consumed across the top, middle, and bottom life expectancy categories. When observing the blue bars, representing the average protein intake from vegetal sources, we see a similar pattern across all categories, with the highest bars for the middle category, followed by the top category, and finally the bottom category, which consumes the least amount of protein from vegetal sources.

However, the most visible difference is in the intake of animal protein. The top countries consume significantly more protein from animal sources, with an average intake between 65 and 68 grams, and this trend is on the rise. The middle-category countries, while still consuming significant vegetal protein, have reached a point where animal protein intake has increased and even surpassed plant-based protein consumption. The bottom-category countries, on the other hand, show the lowest intake of animal protein, consuming only between 15 and 20 grams on average.

These trends lead to important questions. As seen in other plots, the middle-category countries seem to shift toward consuming more animal protein as they develop economically. It aligns with the idea that as countries get wealthier, they can afford more animal products. Conversely, the bottom-category countries simply might not be able to afford animal protein, as it is too expensive for them, which may contribute to their reliance on plant-based protein.

Variations in Animal vs. Plant Protein Ratios Across Countries

Plot 10: Ratio of Animal vs. Plant Based

Similar to Plot 8, Plot 10 illustrates the ratio of animal to plant-based protein intake across the top, middle, and bottom country categories. Interestingly, across all countries, the ratios are relatively volatile, with noticeable variation over the years.

Starting with the top category, we observe that countries like Australia, Sweden, Switzerland, France, and Spain have the highest ratios, meaning that they derive most of their protein from animal sources. However, there seems to be a slight downward trend in the consumption of animal protein in these countries, indicating a shift towards more balanced choice of protein sources, or perhaps a reduction in animal product consumption altogether.

Within the top category, there is a gap between these countries and others, such as Italy, Norway, Korea, Cyprus, and Japan, which tend to consume less animal protein, though their consumption is on the rise.

In the middle category, there is no clear clustering or discernible pattern. The consumption of animal protein appears versatile and fluctuates over the years. Countries like Saint Lucia, for instance, consume a relatively high portion of animal protein, while Bangladesh, at the lower end, consumes as little animal protein as some of the bottom-category countries.

The bottom category displays a wide range of consumption patterns as well. Some countries, like Zimbabwe, consume very low amounts of animal protein, though their intake is slowly rising. Interestingly, some bottom-category countries consume nearly as much animal protein as top-category countries. For instance, the consumption in Zimbabwe is approaching that of the top group, while others, like Niger, remain consistently low in their animal protein consumption.

The variation in animal vs. plant protein consumption across these countries reflects not only economic differences but also shifts in dietary choices and access to different types of protein. The top countries may be choosing to reduce their animal protein intake, while bottom-category countries are still striving to afford more animal products, resulting in an interesting overlap in consumption patterns.

Comparison with Existing Literature on Caloric Intake and Longevity

When comparing these findings with existing literature, some patterns align while others differ. Studies like Ekmekcioglu (2020) suggest that longevity is linked to dietary practices such as caloric restriction and nutrient-dense foods. However, the results here show that countries with the highest life expectancy also tend to consume the most calories, which contrasts with the literature advocating lower caloric intake for longevity.

This suggests factors beyond caloric intake, such as healthcare access and diet quality, may play a key role. For instance, Japan, despite having the highest life expectancy, ranks lower in caloric intake, indicating that a balanced diet and lifestyle may be more important than calorie count alone. Additionally, life expectancy continues to rise in some lower-income countries despite decreasing caloric intake, further challenging the direct link between calories and longevity.